Recovery of a temperate riverine fish assemblage from a major diesel oil spill


Kevin M. Kubach, South Carolina Department of Natural Resources, 311 Natural Resources Drive, Clemson, 29631-3253 SC, U.S.A. E-mail:


1. Biological resilience is of heightened concern in increasingly anthropogenic landscapes. Quantification of faunal resilience across a wide range of spatial scales and geographical areas is necessary to understand factors influencing the rate and degree of recovery, especially in fragmented ecosystems.

2. We evaluated the recovery of a riverine fish assemblage from a major diesel oil pipeline spill and associated fish kill in 37 km of the Reedy River, South Carolina, U.S.A. The fish assemblage was monitored at four disturbed sites within the fish kill zone and one upstream, undisturbed reference site over a 112-month (9.3-year) period following the disturbance. We used non-metric multidimensional scaling (NMS) ordination to evaluate change in fish assemblage structure among sites and to determine the degree of recovery in assemblage structure.

3. NMS ordination of species relative abundance in two dimensions represented 93% of the total variation in fish assemblage structure among samples and illustrated recovery of the fish assemblage. Initial dissimilarity in assemblage structure was evident between the disturbed sites and the reference site, reflecting high mortality from the oil spill. The disturbed sites as a group increased in similarity to the reference assemblage with time, while the reference assemblage remained relatively stable. Strongest similarity in assemblage structure between the disturbed group and the reference group was achieved by October 2000 (52 months post-disturbance), indicating recovery from the oil spill. Remaining variation in assemblage structure was consistent with longitudinal site position and comparable to that of an undisturbed reference river, attributable to inherent longitudinal variation along the 37-kilometre river section.

4. Recovery rate among sites varied in relation to proximity and connectivity to recolonisation sources on a landscape scale. Recovery of the uppermost disturbed site was faster than the other disturbed sites because of its proximity to the undisturbed main stem fish assemblage, whereas the three most downstream sites were slower to recover largely because of isolation by anthropogenic barriers. These observations illustrate the influence of fragmentation on fish assemblage resilience at large spatial scales.